Privacy Policy Compliance of Mobile Sports and Health Apps in China: Scale Development, Data Analysis, and Prospects for Regulatory Reform
Bibliographic record
Abstract
Background: Driven by technological advancements, the proliferation of mobile sports and health apps has revolutionized health management by improving efficiency, cost-effectiveness, and accessibility. While the widespread adoption of these platforms has transformed public health practices and social well-being in China, emerging evidence suggests that inadequacies in their privacy policies may compromise personal information (PI) protection. Objective: This study aimed to conduct a systematic evaluation of privacy policy compliance among 286 mobile sports and health apps in the Chinese Mainland, benchmarking them against the Personal Information Protection Law and associated PI regulatory guidelines. Methods: This study develops a privacy policy compliance indicator scale based on the information life cycle and the legal framework for PI protection in the Chinese Mainland. This scale consists of 5 level 1 indicators and 37 level 2 indicators that assess the privacy policy compliance. Results: The privacy policy compliance of 286 sports and health apps generally performed worse, with only a minimal number (n=11, 3.8%) of apps scoring above 90 points (rated as excellent), nearly half (n=121, 42.3%) of apps scored below 60 points (rated as unqualified). Among the 5 level 1 evaluation indicators for privacy compliance in sports and health apps, the compliance rate for PI collection (mean 74%, SD 25.8%) is the highest, while the compliance rate for PI storage (mean 53.5%, SD 28.4%) is the lowest. The compliance rates for privacy policies across the remaining 3 level 1 evaluation indicators, such as PI usage (mean 54.2%, SD 24.4%), PI entrusted processing, sharing, transferring, and disclosing (mean 62.2%, SD 19.8%), and PI security and feedback (mean 61.7%, SD 21.3%), fall around 60%. Out of 37, 17 level 2 evaluation indicators show a compliance rate below 60%. The compliance rate with privacy policies for 5 level 2 evaluation indicators is exceptionally high, including collection subject (mean 97.2%, SD 16.5%), collection type (mean 99%, SD 10.2%), collection purpose (mean 96.2%, SD 19.3%), reasons for sharing, transferring, and disclosing PI (mean 91.6%, SD 27.8%), and feedback channel (mean 93.4%, SD 24.9%). Notably, 3 indicators exhibit compliance rates below 20%, including sensitive information storage (mean 14%, SD 34.7%), constraints of automatic decision-making (mean 9.4%, SD 29.3%), and deceased user rule (mean 5.2%, SD 22.3%). Authorization for sensitive information (mean 29.4%, SD 45.6%) lagged behind general information (mean 83.6%, SD 37.1%). Conclusions: Although some apps have established commendable policies, there are gaps that compromise the efficacy of PI protection. Considering this, this paper proposes targeted actions for 3 stakeholders: users, regulators, and legislators. Only through coordinated action can the app ecosystem close the compliance gaps, reduce PI protection risks, and restore user trust in digital services.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.053 | 0.063 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.008 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".